JP2022110425A - Medical information management device and medical information management program - Google Patents

Abstract

To allow dynamic information obtained by executing dynamic imaging for a subject including no disease to be used later on.SOLUTION: A medical information management device manages dynamic information obtained by executing dynamic imaging for a subject including no disease and attribute information on the subject having no disease, which are associated with each other. The medical information management device manages dynamic information obtained by executing dynamic imaging for a subject having a disease and attribute information on the subject having a disease, which are associated with each other. The medical information management device determines normality or abnormality of the dynamic information obtained by executing dynamic imaging for a diagnosis object patient by using the above information.SELECTED DRAWING: Figure 3

Description

The present invention relates to a medical information management device, a data structure of medical information, and a medical information management program.

In general, a diagnostic method is used in which abnormalities such as lesions are detected by comparing images obtained by radiography of a patient in the past with current images. However, since images of past normal cases (hereinafter referred to as normal images) are not always available for the patient to be diagnosed, images with similar normal structures to those of the patient to be diagnosed are extracted from a large-scale image database of normal images. Then, a technique has been proposed in which a similar difference image obtained by taking the difference between the extracted normal image and the image of the patient to be diagnosed (hereinafter referred to as the image to be diagnosed) is used as an auxiliary diagnostic image (see Patent Document 1). . However, since there are individual differences in the shape of organs, and artifacts occur in the similarity difference image, a similarity difference image is generated when the shape of the organ matches between the normal image in the image database and the diagnostic target image. I decided to.

By the way, the above-described prior art relates to radiographic still images, and it is known that the maximum expiratory position is used as an example of a normal image in the still images. On the other hand, in recent years, attempts have been made to make diagnoses using dynamic images (dynamic information) obtained from dynamic radiography using radiation. Images are also judged based on knowledge of normal images in still images.

JP 2016-174735 A

However, there are cases where there is no significant difference in the shape of structures such as organs and bones in images of the maximum expiratory position between normal images and images of patients with disease. There is a problem that a doctor cannot sufficiently perform a diagnosis using dynamic images.

In addition, since dynamic images contain much more information than still images, the importance of normal images as diagnostic criteria is increasing.

In addition, since dynamic images are moving images, they require more time to view than still images. Viewing every frame of the video is unacceptable, as it increases the doctor's man-hours for confirmation. Therefore, it is necessary to improve the efficiency of viewing and diagnosing the dynamics of the doctor more than the still images. For this purpose, diagnosis support is required, such as analyzing moving image data and notifying candidates of abnormal points by marking, highlighting, or the like. Since normal images are required for extraction of abnormal site candidates for the purpose of such diagnostic support, collection of normal images is more important for moving images than for still images.

In addition, if an image to be diagnosed does not contain a disease and is erroneously determined to be abnormal, re-imaging of the patient continues, resulting in a considerable increase in radiation exposure compared to re-imaging of a still image. Diagnosis using normal images is also effective in preventing such wasteful radiation exposure.

SUMMARY OF THE INVENTION The present invention has been made in view of the above problems in the prior art, and it is an object of the present invention to make it possible to use dynamic information obtained by performing dynamic imaging of a subject that does not include any disease. do.

In order to solve the above problems, the invention according to claim 1 manages dynamic information obtained by performing dynamic imaging of a subject that does not contain a disease and attribute information of the subject in association with each other. A medical information management device characterized by having management means.

The invention according to claim 2 is characterized in that, in the medical information management apparatus according to claim 1, the disease-free subject is a disease-free subject of a specific tissue.

The invention according to claim 3 is the medical information management apparatus according to claim 2, characterized in that the specific tissue is at least one of respiratory organs, circulatory organs, orthopedic organs, and swallowing related tissues. do.

The invention according to claim 4 is the medical information management apparatus according to any one of claims 1 to 3, wherein the dynamic information includes motion information obtained by analyzing the dynamic image. Characterized by

The invention according to claim 5 is the medical information management apparatus according to claim 4, characterized in that the movement information is information relating to movement of a specific tissue of the subject.

The invention according to claim 6 is the medical information management apparatus according to any one of claims 1 to 5, characterized in that the attribute information does not include information that identifies an individual.

The invention according to claim 7 is the medical information management apparatus according to any one of claims 1 to 6, in which dynamic imaging is performed on the second subject based on the dynamic information. It is characterized by having judgment means for judging whether the second dynamic information is normal or abnormal.

The invention according to claim 8 is the medical information management apparatus according to claim 7, wherein the judgment means judges whether the second dynamic information is normal or abnormal based on the dynamic information and the attribute information. characterized by:

The invention according to claim 9 is the medical information management apparatus according to any one of claims 1 to 6, wherein the management means obtains by performing dynamic imaging of a third subject including a disease. and managing the obtained third motion information and the attribute information of the third subject in association with each other.

According to a tenth aspect of the invention, in the medical information management apparatus of the ninth aspect, the attribute information of the third subject does not include information for identifying an individual.

The invention according to claim 11 is the medical information management device according to claim 9 or 10, wherein the statistical data of the dynamic information, the statistical data of the third dynamic information, and the dynamic data of the second subject It is characterized by having output means for outputting second dynamic information obtained by performing imaging.

The invention according to claim 12 is the medical information management apparatus according to claim 11, wherein the output means outputs the statistical data of the dynamic information, the statistical data of the third dynamic information, and the second dynamic information. It is characterized by outputting on the same screen.

The invention according to claim 13 is the medical information management apparatus according to any one of claims 9 to 12, wherein dynamic imaging of a second subject is performed based on the dynamic information and the third dynamic information. and determining means for determining the normality or abnormality of the second dynamics information obtained by performing the above.

A fourteenth aspect of the present invention is the medical information management apparatus according to the thirteenth aspect, wherein the determination means determines the dynamic information, the attribute information, the third dynamic information, and the third attribute information of the subject. and determining the normality or abnormality of the second dynamic information based on the above.

The invention according to claim 15 is the medical information management device according to any one of claims 7, 8, 13, and 14, wherein the determination means determines whether the second dynamic information is normal or normal by statistical analysis. and determining an abnormality.

The invention according to claim 16 is the medical information management apparatus according to any one of claims 7, 8, 13, and 14, wherein the determination means determines whether the second dynamic information is normal or not by machine learning. and determining an abnormality.

The invention according to claim 17 is the medical information management device according to any one of claims 7, 8, 13, and 14, wherein the normality includes the degree of normality of the second dynamic information, The abnormality is characterized by including the degree of abnormality of the second dynamic information.

According to an eighteenth aspect of the invention, there is provided a data structure of medical information used in a dynamic information processing apparatus for processing dynamic information obtained by performing dynamic imaging, wherein dynamic imaging is performed on a subject that does not include a disease. and the attribute information data of the subject associated with the dynamic information data.

The invention according to claim 19 causes a computer to execute a management process of associating and managing dynamic information obtained by performing dynamic imaging of a subject that does not contain a disease and attribute information of the subject. This is a medical information management program characterized by:

A 20th aspect of the invention is characterized in that, in the medical information management program of the 19th aspect, the disease-free subject is a disease-free subject of a specific tissue.

The invention according to claim 21 is characterized in that, in the medical information management program according to claim 20, the specific organization is at least one of respiratory, circulatory, orthopedic, and swallowing-related tissues. do.

The invention according to claim 22 is the medical information management program according to any one of claims 19 to 21, wherein the dynamic information includes motion information obtained by analyzing dynamic images. Characterized by

The invention according to claim 23 is the medical information management program according to claim 22, characterized in that the movement information is information relating to movement of a specific tissue of the subject.

The invention according to claim 24 is the medical information management program according to any one of claims 19 to 23, characterized in that the attribute information does not include information that identifies an individual.

The invention according to claim 25 is the medical information management program according to any one of claims 19 to 24, wherein the computer performs dynamic imaging of a second subject based on the dynamic information. characterized by executing a judgment process for judging the normality or abnormality of the second dynamics information obtained in .

A twenty-sixth aspect of the present invention is the medical information management program according to the twenty-fifth aspect, wherein in the judgment processing, the normality or abnormality of the second dynamic information is judged based on the dynamic information and the attribute information. characterized by:

The invention according to claim 27 is the medical information management program according to any one of claims 19 to 24, wherein in the management process, a third subject including a disease is captured by dynamic imaging. and managing the obtained third motion information and the attribute information of the third subject in association with each other.

A twenty-eighth aspect of the present invention is the medical information management program according to the twenty-seventh aspect, wherein the attribute information of the third subject does not include information for specifying an individual.

The invention according to claim 29 is the medical information management program according to claim 27 or 28, wherein the computer stores the statistical data of the dynamic information, the statistical data of the third dynamic information, and the second subject. It is characterized by executing an output process of outputting the second dynamic information obtained by performing dynamic imaging on the image.

The invention according to claim 30 is the medical information management program according to claim 29, wherein in the output processing, the statistical data of the dynamic information, the statistical data of the third dynamic information, and the second dynamic information are It is characterized by outputting on the same screen.

The invention according to claim 31 is the medical information management program according to any one of claims 27 to 30, wherein the computer instructs the second subject based on the dynamic information and the third dynamic information. It is characterized by executing a judgment process for judging the normality or abnormality of the second dynamic information obtained by performing the dynamic imaging.

The invention according to claim 32 is the medical information management program according to claim 31, wherein in the determination process, the dynamic information, the attribute information, the third dynamic information, and the attribute information of the third subject are: and determining the normality or abnormality of the second dynamic information based on the above.

The invention according to claim 33 is the medical information management program according to any one of claims 25, 26, 31, and 32, wherein in the judgment processing, the normality or and determining an abnormality.

The invention according to claim 34 is the medical information management program according to any one of claims 25, 26, 31, and 32, wherein in the determination process, the normality or and determining an abnormality.

The invention according to claim 35 is the medical information management program according to any one of claims 25, 26, 31, and 32, wherein the normality includes the degree of normality of the second dynamic information, The abnormality is characterized by including the degree of abnormality of the second dynamic information.

According to the present invention, dynamic information obtained by performing dynamic imaging of a disease-free subject can be used later.

It is a figure showing the whole medical information management system composition in a 1st embodiment of the present invention. It is a block diagram which shows the functional structure of a medical information management apparatus. It is a figure which shows the data structural example of a case database. 4 is a flowchart showing case data registration processing. 4 is a flowchart showing investigation mode processing; It is an example of a search result screen. 4 is a flowchart showing comparison mode processing; It is an example of an analysis result screen. It is an example of an analysis result screen. It is an example of machine learning data used in the second embodiment of the present invention. FIG. 4 is an image diagram showing machine learning processing using patient parameters and diagnosis results. FIG. 4 is an image diagram showing an inference process for predicting a diagnosis prediction result from patient parameters using a learned discriminator; FIG. 10 is a diagram showing processing for creating a normal model by machine learning using patient parameters of normal case data in the third embodiment of the present invention. FIG. 10 is a diagram showing a process of extracting features considered important in determining whether there is an abnormality by machine learning using patient parameters of case data in the fourth embodiment of the present invention.

Hereinafter, embodiments of a medical information management apparatus, a data structure of medical information, and a medical information management program according to the present invention will be described with reference to the drawings. It should be noted that the present invention is not limited to the illustrated examples.

[First embodiment]
(Configuration of medical information management system)
FIG. 1 shows the overall configuration of a medical information management system 100 according to the first embodiment of the invention.
As shown in FIG. 1, the medical information management system 100 includes a medical information management device 10, an imaging console 20, an imaging device 30, an examination device 40, an electronic chart device 50, and a medical image management system 90. is configured with The medical information management apparatus 10, imaging console 20, examination apparatus 40, electronic medical record apparatus 50, and medical image management system 90 are connected via a communication network N such as a LAN (Local Area Network).

The medical information management apparatus 10 has a case database 152, and manages dynamic information such as dynamic images captured by the imaging device 30 in association with attribute information of subjects to be captured.

The imaging console 20 outputs radiation irradiation conditions and image reading conditions to the imaging device 30 to control radiation imaging and radiographic image reading operations by the imaging device 30 .

The imaging device 30 is a device capable of imaging morphological changes in expansion and contraction of the lungs associated with respiratory motion, heartbeats, and other dynamics. In dynamic imaging, the subject is irradiated with pulses of radiation such as X-rays repeatedly at predetermined time intervals (pulse irradiation), or low doses of radiation are continuously irradiated without interruption (continuous irradiation). , acquire a plurality of images showing the dynamics of the subject. A series of images obtained by dynamic imaging are called dynamic images. Also, each of the plurality of images forming the dynamic image is called a frame image. A case where dynamic imaging is performed by pulse irradiation will be described below as an example.
Dynamic photography includes moving image photography, but does not include still image photography while displaying a moving image. Also, dynamic images include moving images, but do not include images obtained by capturing still images while displaying moving images.

The imaging device 30 includes a radiation source 31 , a radiation irradiation control device 32 , a radiation detection section 33 and a reading control device 34 .
The radiation source 31 is arranged at a position facing the radiation detection unit 33 with the subject interposed therebetween, and irradiates the subject with radiation according to the control of the radiation irradiation control device 32 .

The radiation irradiation control device 32 is connected to the imaging console 20 and controls the radiation source 31 based on the radiation irradiation conditions input from the imaging console 20 to perform radiation imaging. Radiation irradiation conditions include, for example, pulse rate, pulse width, pulse interval, number of imaging frames per imaging, tube current, tube voltage, additional filter type, and the like. Pulse rate is the number of radiation shots per second. The pulse width is the radiation exposure time per radiation exposure. The pulse interval is the time from the start of one radiation exposure to the start of the next radiation exposure.

The radiation detection unit 33 is composed of a semiconductor image sensor such as an FPD (Flat Panel Detector). The FPD has a glass substrate or the like, and detects the radiation emitted from the radiation source 31 at a predetermined position on the substrate and transmitted through at least the object according to its intensity, and converts the detected radiation into an electrical signal. A plurality of detection elements (pixels) are arranged in a matrix. Each pixel includes a switching section such as a TFT (Thin Film Transistor).

The reading control device 34 is connected to the imaging console 20, and controls the switching unit of each pixel of the radiation detection unit 33 based on the image reading conditions input from the imaging console 20, and accumulates in each pixel. The image data is obtained by switching the reading of the electric signals received and reading the electric signals accumulated in the radiation detection section 33 . This image data is each frame image of the dynamic image. The reading control device 34 outputs the acquired dynamic image to the imaging console 20 . The image reading conditions include, for example, frame rate, frame interval, pixel size, image size (matrix size), and the like. The frame rate is the number of frame images acquired per second and matches the pulse rate. The frame interval is the time from the start of the acquisition operation of one frame image to the start of the acquisition operation of the next frame image, and coincides with the pulse interval.

The testing device 40 is a device that performs a respiratory function test (PFT: Pulmonary Function Test) on a patient. Pulmonary function tests include vital capacity (VC), total lung capacity (TLC), functional residual capacity (FRC), residual capacity (RV), RV/TLC, expiratory reserve volume (ERV), expiratory volume in 1 second (FEV1 ), etc. are measured.

The electronic chart device 50 manages chart information for each patient. The chart information includes patient information about the patient.

The medical image management system 90 manages image information for each patient. A medical image management system 90 contains image information about a patient.

(Configuration of medical information management device)
FIG. 2 shows the functional configuration of the medical information management device 10. As shown in FIG.
As shown in FIG. 2, the medical information management apparatus 10 includes a control section 11, an operation section 12, a display section 13, a communication section 14, and a storage section 15, which are connected by a bus.

The control unit 11 includes a CPU (Central Processing Unit), a RAM (Random Access Memory), and the like. The CPU of the control unit 11 reads the system program and various processing programs stored in the storage unit 15, expands them in the RAM, and executes various processes according to the expanded programs.

The operation unit 12 includes a keyboard having cursor keys, character/number input keys, various function keys, etc., and a pointing device such as a mouse, and controls instruction signals input by key operations on the keyboard and mouse operations. Output to unit 11 . Moreover, the operation unit 12 may include a touch panel on the display screen of the display unit 13 , and in this case, outputs an instruction signal input through the touch panel to the control unit 11 .

The display unit 13 is configured by a monitor such as an LCD (Liquid Crystal Display), and displays various screens according to instructions of display signals input from the control unit 11 .

The communication unit 14 includes a LAN adapter, a modem, a TA (Terminal Adapter), etc., and controls data transmission/reception with each device connected to the communication network N. FIG.

The storage unit 15 is configured by a nonvolatile semiconductor memory, hard disk, or the like. The storage unit 15 stores various programs including the medical information management program 151 executed by the control unit 11, parameters necessary for execution of processing by the programs, and data such as processing results. A case database 152 is stored in the storage unit 15 .

FIG. 3 shows a data configuration example of the case database 152. As shown in FIG.
As shown in FIG. 3, the case database 152 stores dynamic information, attribute information, and diagnosis results in association with each other. In the case database 152, one line of data is treated as one case data.

Dynamic information is information obtained by performing dynamic imaging of a subject, and includes at least one of a dynamic image and motion information.
A dynamic image is a series of images (image data) obtained by dynamic imaging. In the case database 152, a path name indicating the storage location of the dynamic image file may be associated instead of the dynamic image.
The motion information is information representing the motion of the subject's tissues (for example, respiratory, circulatory, orthopedic, and swallowing-related tissues) obtained from the dynamic image. The motion information includes information such as the position obtained for each frame image, the speed obtained from the difference between the frame images, the maximum speed obtained analytically from the information, the change rate of the size, and the like. As the movement information, for example, information quantifying the movement of tissues such as diaphragm velocity, lung area change rate, airway diameter constriction rate, and the like is used. Also, when the motion information is time-series data, it may be graphed over time.

When the posterior ribs, sternum, clavicle, spine, diaphragm, and ribcage are targeted for imaging, time-series changes in position, time-series changes in velocity, maximum distance from the initial position, maximum and minimum velocities, etc. are used as motion information. be done.
When the heart is an imaging target region, time-series change in size, time-series change in signal value density, size change rate, signal value density change rate, and the like are used as motion information.
When the aortic arch is the imaging target region, time-series changes in signal value density, rate of change in signal value density, and the like are used as motion information.
When the trachea is a part to be imaged, time-series changes in tracheal diameter size, tracheal diameter narrowing ratio, and the like are used as motion information.
When the lung field is the part to be imaged, time-series change in lung field size, maximum/minimum lung field area change rate, signal value density change rate, and the like are used as motion information.

The attribute information is information indicating the attributes of the subject that was taken as the shooting target of the dynamic image. Attribute information includes, for example, age, gender, height, weight, BMI, race, smoking history, and the like. Attribute information does not include information that identifies an individual. Information that identifies an individual includes name, address, telephone number, and the like.
Moreover, it is good also as using test results, such as a respiratory function test result, as attribute information. Pulmonary function test results include vital capacity (VC), total lung capacity (TLC), functional residual capacity (FRC), residual capacity (RV), RV/TLC, expiratory reserve volume (ERV), F1 second (FEV1), etc. are included.
Also, as the attribute information, imaging conditions (radiation irradiation conditions, image reading conditions, etc.) at the time of imaging the dynamic image in the imaging device 30 may be used. The imaging conditions may include imaging date/time, body part, imaging direction, and the like.

The diagnosis result is a diagnosis result for the dynamic information (object to be photographed for the dynamic image), and includes a normal/abnormal flag and a diagnosis name.
The normal/abnormal flag indicates whether or not the subject has a disease, and is "normal" if the subject does not have a disease, and "abnormal" if the subject has a disease.
The diagnosis name is a diagnosis name (disease name, etc.) when the normal/abnormal flag is "abnormal", that is, when the subject includes a disease. If the normal/abnormal flag is "normal", the diagnosis name is "no disease". Note that the diagnosis name may be blank when the normal/abnormal flag is "normal".
Diseases include, for example, respiratory, cardiovascular, orthopedic, and swallowing diseases. More specifically, respiratory diseases include COPD (chronic obstructive pulmonary disease) and pneumonia, cardiovascular diseases include heart failure and pulmonary embolism, and orthopedic diseases include arthrosis and bone fractures.
In addition, normal/abnormal flags are not only normal/abnormal flags for all diseases, but also normal/abnormal flags for specific diseases, such as a normal/abnormal flag for respiratory diseases and a normal/abnormal flag for cardiovascular diseases. may

In the case database 152, a record with a normal/abnormal flag of "normal" is a normal case, and a record with a normal/abnormal flag of "abnormal" is an abnormal case.

The case database 152 has a data structure of medical information used by the medical information management apparatus 10 as a dynamic information processing apparatus for processing dynamic information obtained by performing dynamic imaging. Here, the dynamic information to be processed is a dynamic image or motion information of a patient to be diagnosed. Processing of dynamic information includes not only image processing of dynamic images of a patient to be diagnosed, but also display processing (simultaneous display with normal case data, etc.), statistical analysis, and inference (machine learning, deep learning).

Statistical analysis is the process of performing statistical analysis on the data to be processed, such as the process of calculating the average value, median value, standard deviation, etc., and the process of analyzing the data distribution by creating a histogram. including.

Inference is the process of deriving an inference result from data to be processed based on the results of machine learning. The use of case data managed by the case database 152 for machine learning will be described in the second embodiment and thereafter.

By accessing the case database 152, the user can use the case data managed by the case database 152. FIG. For example, when it is desired to analyze movement information of the diaphragm (diaphragm position, speed, etc.) by designating age and gender, case data is extracted from the case database 152 by designating the age range and gender as search conditions. Then, the number of data (number of extracted cases), average age (average age of extracted case data), average amplitude (average diaphragm amplitude of extracted case data), average velocity, etc. (extracted Diaphragm velocity average value of case data) is calculated, and the analysis result is displayed, stored in a recording medium, or transmitted to an external device.

The control unit 11 may include a disease (for example, a fracture) in a subject that does not include a disease (for example, when dynamic information is obtained by dynamic imaging of tissues related to the respiratory system) other than the respiratory system (for example, an arm that is orthopedic, etc.) ) obtained by performing dynamic imaging (hereinafter referred to as "normal dynamic information") and the attribute information of the subject that does not include disease (hereinafter referred to as "normal attribute information") Associate and manage. That is, the control unit 11 functions as management means.
The normal dynamics information includes both cases in which the subject was photographed after being known to have no disease in advance, and cases in which the subject was diagnosed not to have a disease based on the dynamics information after imaging.

The control unit 11 (managing means) further provides third dynamic information (hereinafter referred to as “abnormal dynamic information”) obtained by performing dynamic imaging of a third subject including a disease, and the disease. Attribute information of the third subject (hereinafter referred to as "abnormal attribute information") is associated with the attribute information and managed.
The abnormal dynamics information includes both cases in which the subject was photographed after it was known in advance that the subject had a disease, and cases in which the subject was diagnosed as having a disease based on the dynamics information after imaging.

Based on the normal dynamics information, the control unit 11 obtains second dynamics information (hereinafter referred to as "diagnostic dynamics Information”) is normal or abnormal. That is, the control unit 11 functions as determination means.

Normality is information indicating whether or not the diagnostic target dynamic information is normal, and includes the degree of normality of the diagnostic target dynamic information.
Abnormality is information indicating whether or not the diagnostic target dynamic information is abnormal, and includes the degree of abnormality of the diagnostic target dynamic information.

For example, the control unit 11 compares the normal dynamics information and the diagnostic target dynamics information, and determines that the diagnostic target dynamics information is normal when the diagnostic target dynamics information is close to the normal dynamics information (there is no difference). . On the other hand, when the diagnostic target dynamic information is significantly different from the normal dynamic information, the control unit 11 determines that the diagnostic target dynamic information is abnormal. A predetermined threshold or the like can be used to determine whether the diagnostic target dynamic information and the normal dynamic information are similar or different.

The control unit 11 (judgment means) judges the normality or abnormality of the diagnosis target dynamics information based on the normal dynamics information and the normal attribute information. Specifically, the control unit 11 compares the diagnostic target dynamics information with the normal dynamics information corresponding to the same or similar attribute information (normal attribute information) as the diagnostic target patient's attribute information, thereby determining the diagnosis target dynamics. Judge the normality or abnormality of information. A predetermined threshold or the like can be used to determine similarity in attribute information.

Based on the normal dynamics information and the abnormal dynamics information, the control unit 11 (judgment means) judges the normality or abnormality of the diagnostic target dynamics information obtained by performing dynamic imaging of the diagnosis target patient.

For example, the control unit 11 compares the normal dynamics information and the abnormal dynamics information with the diagnostic dynamics information, and if the diagnostic dynamics information is closer to the normal dynamics information than the abnormal dynamics information, the diagnostic dynamics information is normal. judge there is. On the other hand, the control unit 11 determines that the diagnostic target dynamic information is abnormal when the diagnostic target dynamic information is closer to the abnormal dynamic information than the normal dynamic information, or when it is far from the normal dynamic information. A predetermined threshold or the like can be used to determine whether the diagnostic target dynamic information and the normal dynamic information or abnormal dynamic information are close or apart.

The control unit 11 (judgment means) judges the normality or abnormality of the diagnosis target dynamics information based on the normal dynamics information, the normal attribute information, the abnormal dynamics information, and the abnormal attribute information. Specifically, the control unit 11 converts normal dynamics information and abnormal dynamics information corresponding to attribute information (normal attribute information, abnormal attribute information) identical or similar to attribute information of a patient to be diagnosed, and diagnostic target dynamics information. By comparing, the normality or abnormality of the diagnosis target dynamic information is determined.

The control unit 11 (judgment means) judges the normality or abnormality of the diagnosis target dynamic information by statistical analysis. The distribution of motion information (histogram, etc.), average value, median value, standard deviation, etc. can be used for statistical analysis.

The control unit 11 outputs statistical data of normal dynamics information, statistical data of abnormal dynamics information, and diagnostic target dynamics information obtained by performing dynamic imaging on a patient to be diagnosed. That is, the control unit 11 functions as output means.
The control unit 11 (output means) outputs the statistical data of the normal dynamics information, the statistical data of the abnormal dynamics information, and the diagnostic target dynamics information on the same screen.

(Operation of medical information management device)
Next, the operation of the medical information management apparatus 10 will be described.
FIG. 4 is a flow chart showing case data registration processing executed by the medical information management apparatus 10 . The case data registration process is a process of collecting and registering data of normal cases and abnormal cases. Realized.

First, when dynamic imaging is performed on a subject in the imaging device 30, the control unit 11 of the medical information management device 10 acquires a dynamic image from the imaging console 20 via the communication unit 14 (step S1). . A dynamic image is composed of a plurality of frame images (for example, 15 frames per second).

Next, the control unit 11 acquires attribute information corresponding to the subject that was taken as a shooting target in the dynamic image (step S2). Specifically, the control unit 11 causes the display unit 13 to display an attribute information input screen, and accepts an input of attribute information by a user's operation from the operation unit 12 .
Note that the control unit 11 may acquire the attribute information from additional information attached to the moving image file, or the patient information (attribute information) corresponding to the subject from the electronic medical chart device 50 via the communication unit 14 . ) may be obtained. Here, if the acquired information includes information identifying an individual such as the patient's name, the control unit 11 excludes this information.
In addition, the control unit 11 may acquire the result of a respiratory function test performed on the subject (patient) from the testing device 40 via the communication unit 14 as the attribute information corresponding to the subject.

Next, the control unit 11 analyzes the dynamic image acquired in step S1 and calculates motion information (step S3). The motion information includes respiratory, circulatory, shaping, and swallowing motion information. For example, respiratory motion information includes diaphragm velocity, lung area change rate, and airway diameter constriction rate. The motion information about the circulatory system also includes the speed at which the heart wall moves. In addition, the motion information related to shaping includes the trajectory of bending and stretching of joints such as knees and elbows (information on position change) and the speed of stretching the joints. For example, the control unit 11 detects the position of the diaphragm from dynamic images (a plurality of frame images) obtained by photographing the front of the chest, and calculates the diaphragm velocity between the frame images. Further, the control unit 11 obtains the diaphragm maximum velocity from the movement of the diaphragm in a series of dynamic images.
Further, the control unit 11 detects the position (region) of the lung field from the dynamic image of the chest front, and calculates the lung field area for each frame image. Then, the control unit 11 calculates the lung field area change rate from the movement of the lung field in a series of dynamic images.
Further, the control unit 11 detects the position of the airway from the dynamic image of the front of the chest and calculates the diameter of the airway for each frame image. Then, the control unit 11 calculates the airway diameter constriction rate from the movement of the airway in a series of dynamic images.

Next, the control unit 11 determines whether or not the dynamic image acquired in step S1 includes a disease (step S4). Specifically, the control unit 11 causes the display unit 13 to display a diagnosis result input screen, and allows the user to operate the operation unit 12 to input the presence or absence of a disease (normal/abnormal), the diagnosis name (including the disease), (if any).

If the dynamic image does not include a disease (step S4; NO), the control unit 11 associates the dynamic information (dynamic image, motion information) with the attribute information, and stores them in the case database 152 ( Step S5), the diagnosis name corresponding to this case is set to "no disease". Further, the control unit 11 sets the normal/abnormal flag corresponding to this case to "normal" (step S6).

In step S4, if the dynamic image includes a disease (step S4; YES), the control unit 11 generates dynamic information (dynamic image, motion information), attribute information, and diagnosis name (diagnosis result). are associated with each other and stored in the case database 152 (step S7). Further, the control unit 11 sets the normal/abnormal flag corresponding to this case to "abnormal" (step S8).
After step S6 or step S8, the case data registration process ends.

In the case data registration process, the dynamic image is acquired from the imaging console 20 at the timing when the dynamic imaging is performed by the imaging device 30 (step S1), but the acquisition timing of the dynamic image is limited to this. It is also possible to obtain a dynamic image that has been captured and accumulated in advance instead of the original.

FIG. 5 is a flowchart showing investigation mode processing executed by the medical information management apparatus 10. As shown in FIG. Investigation mode processing is processing for retrieving and referring to case data accumulated in the case database 152 irrespective of cases to be diagnosed. It is realized by software processing in collaboration with

First, the control unit 11 causes the display unit 13 to display an input screen for search conditions, and accepts input of search conditions by user's operation from the operation unit 12 (step S11). Specifically, "item", "attribute range", and "normal/abnormal" are input as search conditions.
"Item" is an item of motion information to be processed.
The “attribute range” is a search condition related to attributes for narrowing down the cases to be searched.
"Normal/abnormal" is a search condition indicating whether the search target is to search only normal cases, only abnormal cases, or normal cases and abnormal cases.

Next, the control unit 11 extracts data matching the search conditions from the case database 152 (step S12).
For example, if the search target is limited to only normal cases, the control unit 11 determines from the case database 152 that the "normal/abnormal flag" is "normal" and the "attribute information" is Extract case data included in the entered attribute range.
If the search target is set to be a normal case or an abnormal case, the control unit 11 retrieves from the case database 152, regardless of the "normal/abnormal flag", the "attribute information" within the attribute range input in step S11. Extract the included case data.

Next, the control unit 11 graphs the processing target items based on the extracted case data (step S13). For example, the control unit 11 creates a histogram in which the horizontal axis represents the classes obtained by classifying the numerical value of the item to be processed into a plurality of classes, and the vertical axis represents the number of cases (frequency). When the search targets are set to be normal cases and abnormal cases, the normal cases and abnormal cases are graphed separately, for example, by using different colors.

Next, the control unit 11 causes the display unit 13 to display the search result (step S14).
With this, the investigation mode processing is completed.

FIG. 6 shows an example of a search result screen 131 displayed on the display unit 13. As shown in FIG.
The search result screen 131 includes a search condition display area 131A and a search result display area 131B.

A mode column 60, an item column 61, and an option column 62 are provided in the search condition display area 131A.
A mode column 60 displays the mode selected by the user. In the search condition display area 131A, the mode column 60 displays "investigation mode".
The item column 61 displays items input by the user as search conditions. In the search condition display area 131A, the item column 61 displays "diaphragm maximum velocity".
The option column 62 displays search conditions (option conditions) other than the items entered by the user. In the search condition display area 131A, the option column 62 displays "40's and 50's" and "Normal". Here, normal cases in their 40s or 50s are searched.

The search result display area 131B includes a total data number column 70, an average age column 71, an average value column 72, a median value column 73, a graph display column 80, and the like.
The total data count column 70 displays the total data count of cases that match the search conditions.
The average age column 71 displays the average age of cases that match the search conditions. Numerical values in parentheses in the average age column 71 are the minimum and maximum ages of cases that match the search conditions. An average value column 72 and a median value column 73 display the average value and median value of the processing target item (diaphragmatic maximum velocity) of cases that match the search conditions, respectively. Numerical values in parentheses in the mean value column 72 and the median value column 73 are the minimum and maximum values in the processing target item (maximum diaphragm velocity) of cases that match the search conditions.

The graph display field 80 displays a histogram of the processing target item (diaphragmatic maximum velocity) of the cases that match the search conditions. In the search result display area 131B, the graph display column 80 displays the distribution of the "maximum diaphragm velocity" of normal cases in their 40's or 50's.

Further, an abnormal case addition instruction section 74 and an item condition change instruction section 75 are provided in the search result display area 131B. The abnormal case addition instruction section 74 and the item condition change instruction section 75 are links to related information.
When the user operates the operation unit 12 to press the abnormal case addition instruction unit 74, the search results including the abnormal cases are displayed in the search result display area 131B.
When the user operates the operation unit 12 to press the item condition change instruction unit 75, the search result display area 131B displays the search results in which the processing target item is changed to "lung area change rate". .

FIG. 7 is a flowchart showing comparison mode processing executed by the medical information management apparatus 10. As shown in FIG. The comparison mode process is a process for judging the normality/abnormality of a diagnosis target case by comparing with the case data accumulated in the case database 152, and is stored in the CPU of the control unit 11 and the storage unit 15. It is realized by software processing in collaboration with the medical information management program 151 .

First, the control unit 11 causes the display unit 13 to display a search condition input screen, and receives an input of search conditions by the user's operation from the operation unit 12 (step S21). Specifically, "item", "attribute range", and "normal/abnormal" are input as search conditions. The search conditions are the same as in step S11 of the investigation mode process (see FIG. 5).
As for the "attribute range", a range including the attributes of the diagnosis target case may be specified.

Next, the control unit 11 causes the display unit 13 to display a screen for specifying the dynamic information related to the diagnosis target case, and the user operates the operation unit 12 to specify the diagnostic target dynamic information (dynamic image or motion information). Accept (step S22). The user may specify the dynamic image itself as the diagnosis target dynamic information, or may specify the motion information (numerical value, etc.) corresponding to the processing target item (item input as the search condition). Diagnosis target dynamic information may be prepared in advance in the storage unit 15 of the medical information management apparatus 10 or an external device, or a dynamic image captured by the imaging device 30 may be acquired from the imaging console 20.

Here, the control unit 11 determines whether or not the designated diagnostic object dynamic information is a dynamic image (step S23).

If the designated diagnostic target dynamic information is a dynamic image (step S23; YES), the control unit 11 analyzes the diagnostic target dynamic image and calculates a value corresponding to the processing target item (step S24). .

After step S24 or in step S23, if the specified diagnostic target dynamic information is not a dynamic image (step S23; NO), that is, the motion information corresponding to the processing target item is specified as the diagnostic target dynamic information. In this case, the control unit 11 extracts data that matches the search condition from the case database 152 (step S25).
For example, if the search target is limited to only normal cases, the control unit 11 determines from the case database 152 that the "normal/abnormal flag" is "normal" and the "attribute information" is Extract case data included in the entered attribute range.
When the search target is set to normal cases and abnormal cases, the control unit 11 retrieves from the case database 152, regardless of the "normal/abnormal flag", the "attribute information" within the attribute range input in step S21. Extract the included case data.

Next, the control unit 11 judges the normality/abnormality of the dynamic information to be diagnosed (step S26).
For example, when the search target is limited to normal cases only, the control unit 11 determines the normality/abnormality of the diagnostic target dynamic information based on the normal case data extracted from the case database 152 .
When the search target is set to normal cases and abnormal cases, the control unit 11 determines the normality/abnormality of the diagnostic target dynamic information based on the normal case data and the abnormal case data extracted from the case database 152. to judge.

Judgment of normality/abnormality of diagnostic target dynamic information may be made as normal (not including disease) or abnormal (including disease), or degree of normality (probability, possibility, etc.) or degree of abnormality (probability). , possibility, etc.). In addition to obtaining the degree of normality or abnormality by a numerical value or level, generating information that enables recognition of where the dynamic information to be diagnosed is positioned in the distribution of normal cases and/or abnormal cases. It's okay. Further, the control unit 11 may specify a diagnosis name when it is determined that there is an abnormality (including a disease).

Next, the control unit 11 graphs the processing target items based on the extracted case data (step S27). For example, the control unit 11 creates a histogram in which the horizontal axis represents the classes obtained by classifying the numerical value of the item to be processed into a plurality of classes, and the vertical axis represents the number of cases (frequency). When the search targets are set to be normal cases and abnormal cases, the normal cases and abnormal cases are graphed separately, for example, by using different colors.

Next, the control unit 11 adds the position of the diagnostic target dynamic information to the graph (step S28). For example, the control unit 11 attaches a mark or the like to a position on the graph where the class to which the diagnosis target dynamic information belongs can be recognized.

Next, the control unit 11 causes the display unit 13 to display the analysis result (step S29). For example, the control unit 11 stores statistical data of normal dynamics information (when normal cases are included in the search target), statistical data of abnormal dynamics information (when abnormal cases are included in the search target), diagnostic target dynamics Information is displayed on the same screen of the display unit 13 .
With this, the comparison mode processing ends.

FIG. 8 shows an example of the analysis result screen 132 displayed on the display unit 13. As shown in FIG. The analysis result screen 132 is an example in which movement information (diaphragm maximum velocity value) is specified as diagnostic target dynamic information.
The analysis result screen 132 includes a search condition display area 132A and an analysis result display area 132B.

A mode column 60, an item column 61, an option column 62, and a diagnosis target value column 63 are provided in the search condition display area 132A.
The mode column 60, item column 61, and option column 62 are the same as those of the search result screen 131 (see FIG. 6).
In the search condition display area 132A, the mode column 60 displays "comparison mode".
In the search condition display area 132A, the item field 61 displays "diaphragm maximum velocity".
In the search condition display area 132A, the option column 62 displays "40's and 50's" and "normal/abnormal". Here, normal cases and abnormal cases in their 40s or 50s are searched.
In the diagnosis target value column 63, movement information (value of a processing target item) specified by the user is displayed as diagnostic target dynamic information. In the search condition display area 132A, the diagnosis target value column 63 displays a value of "42.1" as the maximum diaphragm velocity.

The analysis result display area 132B includes a total data number column 70, an average age column 71, an average value column 72, a median value column 73, a normal/abnormal judgment result column 76, a diagnosis target value column 77, a graph display column 80, and the like. .
The total data number column 70, the average age column 71, the average value column 72, and the median value column 73 are the same as those of the search result screen 131 (see FIG. 6).

The normality/abnormality determination result column 76 displays the normality or abnormality determined for the diagnosis target value (the value of the processing target item specified as the diagnostic target dynamic information). Here, the normality/abnormality judgment result column 76 displays a normality probability of "86.1%" as the degree of normality. For example, the ratio (%) of the number of normal cases to the total number of cases (number of normal cases + number of abnormal cases) included in the same class as the class to which the diagnosis target value belongs is defined as the probability of being normal.

Similar to the diagnostic target value column 63, the diagnostic target value column 77 displays the movement information (the value of the processing target item) specified by the user as the diagnostic target dynamic information.

The graph display field 80 displays a histogram of the processing target items of the cases that match the search conditions. In the analysis result display area 132B, the graph display field 80 displays the distribution of the "maximum diaphragm velocity" for cases in their 40s or 50s, divided into normal and abnormal.
Further, in the graph display field 80, a star mark 81 indicates to which class on the histogram the value "42.1" specified as the diagnosis target value belongs.

On the analysis result screen 132, statistical data of normal dynamics information, statistical data of abnormal dynamics information, and diagnostic target dynamics information are output on the same screen.
In the analysis result screen 132 , the statistical data of the normal dynamics information corresponds to the histogram of normal cases in the graph display field 80 . Outputs (displays) the total number of normal case data among the case data that matches the search conditions, and the average value and median value of the processing target items calculated only from the normal case data, as statistical data of normal dynamics information. You can do it.
On the analysis result screen 132 , the statistical data of the abnormal dynamics information corresponds to the histogram of abnormal cases in the graph display field 80 . Outputs (displays) the total number of abnormal case data among the case data that matches the search conditions, and the average value and median value of the processing target items calculated only from the abnormal case data, as statistical data of the abnormal dynamics information. You can do it.
On the analysis result screen 132 , the diagnostic target dynamic information corresponds to the value “42.1” in the diagnostic target value column 77 and the star mark 81 in the graph display column 80 .

FIG. 9 shows an example of the analysis result screen 133 displayed on the display unit 13. As shown in FIG. The analysis result screen 133 is an example when a dynamic image is designated as diagnostic target dynamic information.
The analysis result screen 133 has almost the same configuration as the analysis result screen 132 (see FIG. 8). Only different parts will be explained.
The analysis result screen 133 includes a search condition display area 133A and an analysis result display area 133B.

In place of the diagnostic target value column 63 of the search condition display region 132A (see FIG. 8), a diagnostic target image input presence/absence column 64 and a diagnostic target image column 65 are provided in the search condition display region 133A.
The input/non-existence of diagnostic target image input field 64 displays whether or not a dynamic image is input as the diagnostic target dynamic information.
In the diagnostic target image column 65, the dynamic image specified by the user is displayed when the diagnostic target image input presence/absence column 64 is "Yes".

A calculated value column 78 is provided in the analysis result display area 133B instead of the diagnosis target value column 77 of the analysis result display area 132B (see FIG. 8).
The calculated value column 78 displays the value of the processing target item "42.1" calculated from the input image. Even if the value of the processing target item for the diagnosis target is not calculated, the comparison search can be performed by inputting the image.
The normality/abnormality determination result column 76 displays the normality or abnormality determined for the value of the processing target item calculated from the input image.

In the graph display column 80, a star mark 81 indicates to which class on the histogram the value of the processing target item calculated from the input image belongs.

Also on the analysis result screen 133, statistical data of normal dynamics information (histogram of normal cases in graph display column 80), statistical data of abnormal dynamics information (histogram of abnormal cases in graph display column 80), and diagnostic target dynamics information ( The value in the calculated value column 78, the star mark 81 in the graph display column 80, and the dynamic image in the diagnosis target image column 65) are output on the same screen.

As described above, according to the first embodiment, in the medical information management apparatus 10, dynamic information (normal dynamic information) obtained by performing dynamic imaging of a subject that does not include disease, and By managing the attribute information (normal attribute information) of the subject that does not include the disease, the dynamic information obtained by performing dynamic imaging on the subject that does not include the disease can be used later. This can contribute to highly accurate diagnosis support and actual clinical practice. It can also be used for education and research of medical workers.

In addition, since the attribute information (normal attribute information, abnormal attribute information) does not include information that identifies an individual, it is possible to prevent leakage of personal information of the subject (patient) targeted for dynamic imaging.

In addition, since the normal dynamics information is accumulated, the second dynamics information (diagnostic target dynamics information) obtained by performing dynamic imaging of the second subject (diagnosis target patient) based on the normal dynamics information ) can be judged normal or abnormal.
In addition, since the normal dynamics information and the normal attribute information are associated and accumulated, the normality or abnormality of the diagnostic target dynamics information can be determined based on the normal dynamics information and the normal attribute information.

Further, in the medical information management apparatus 10, in addition to the normal dynamics information and the normal attribute information, third dynamics information (abnormal dynamics information) obtained by performing dynamic imaging of a third subject including a disease, By managing the attribute information (abnormal attribute information) of the third subject including diseases in association with each other, the normal behavior information and the abnormal behavior information can be used for future diagnosis and the like.

In addition, since the normal dynamics information and the abnormal dynamics information are accumulated, the normality or the abnormal dynamics information of the diagnostic target dynamics obtained by performing dynamic imaging of the diagnosis target patient based on the normal dynamics information and the abnormal dynamics information Anomalies can be determined.
In addition, since the normal dynamics information and the normal attribute information and the abnormal dynamics information and the abnormal attribute information are associated and accumulated, the diagnostic target dynamics can be determined based on the normal dynamics information, the normal attribute information, the abnormal dynamics information and the abnormal attribute information. It is possible to judge whether the information is normal or abnormal.

In addition, when the user designates an item to be processed (motion information) or attribute, only relevant information can be extracted from the case database 152 and displayed statistically.
For example, as shown in the analysis result screen 132 of FIG. 8 and the analysis result screen 133 of FIG. 9, statistical data of normal dynamics information, statistical data of abnormal dynamics information, and diagnostic target dynamics information can be output. By distinguishingly displaying normal cases and abnormal cases or showing the normal range and/or the abnormal range on a graph, it is possible to provide the difference in dynamics between healthy subjects and abnormal subjects.

If the attribute information is not used for determining the normality or abnormality of the dynamic information to be diagnosed, limitation by the attribute information (input of the attribute range in step S21) is eliminated in the comparison mode processing (see FIG. 7). , case data of all attributes in the case database 152 may be used for comparison.

Also, the extracted statistical data can be output to the outside. Data formats for output include, for example, csv and pdf formats.

[Second embodiment]
Next, a second embodiment to which the present invention is applied will be described.
The medical information management system according to the second embodiment has the same configuration as the medical information management system 100 shown in the first embodiment, so illustration and description of the configuration are omitted. The configuration and processing that are characteristic of the second embodiment will be described below.

The medical information management apparatus 10 utilizes artificial intelligence (AI) to obtain second dynamic information (diagnostic target dynamic information) is normal or abnormal.

The control unit 11 of the medical information management apparatus 10 provides dynamic information (normal dynamic information) obtained by performing dynamic imaging of a subject that does not include disease, attribute information of the subject that does not include disease (normal attribute information), Based on the third dynamic information (abnormal dynamic information) obtained by performing dynamic imaging of the third subject including the disease and the attribute information (abnormal attribute information) of the third subject including the disease, Judge the normality or abnormality of diagnostic target dynamic information.

Specifically, based on the case data accumulated in the case database 152 (see FIG. 3), the control unit 11 performs machine learning for determining the normality or abnormality of the dynamic information, and based on the learning result, Then, the normality or abnormality of the dynamic information to be diagnosed is determined. The machine learning and the use of the learning results are realized by software processing through cooperation between the CPU of the control unit 11 and the medical information management program 151 stored in the storage unit 15 .

In the medical information management apparatus 10, the case database 152 (see FIG. 3) associates dynamic information, attribute information, and diagnosis results. It can be said that the patient parameter to be obtained is associated with the diagnosis result. Using this correspondence relationship, the control unit 11 performs machine learning with patient parameters as input and diagnosis results as output, and generates a discriminator for determining normality or abnormality of diagnostic target dynamic information. The control unit 11 uses a learned discriminator to determine the normality or abnormality of the diagnostic target dynamic information.

As machine learning, Support Vector Machine (SVM), Random Forest, Deep Learning, etc. can be used.

FIG. 11 is an image diagram showing machine learning processing using patient parameters (input data) and diagnosis results (correct data). The control unit 11 stores age (attribute information), gender (attribute information), smoking history (attribute information), height (attribute information), weight (attribute information), BMI (attribute information), Patient parameters such as respiratory function test results (attribute information), lung area change rate (dynamic information), airway diameter narrowing rate (dynamic information), diaphragmatic velocity (dynamic information), etc. are input, and no disease (normal), COPD, A discriminator is generated by using diagnostic results of bronchial asthma, lung cancer, diabetes, etc. as an output.

FIG. 12 is an image diagram showing inference processing for predicting diagnostic prediction results (prediction data) from patient parameters (input data) to be diagnosed using a learned classifier. The control unit 11 provides attribute information (age, sex, smoking history, height, weight, BMI, respiratory function test results, etc.) and dynamic information (lung field area change rate, airway diameter constriction rate, diaphragmatic velocity, etc.) of the patient to be diagnosed. is input to the learned discriminator, and an output result (diagnostic prediction result) is obtained. The control unit 11 outputs no disease (normal) or a diagnosis (COPD, bronchial asthma, lung cancer, diabetes, etc.) as a diagnosis prediction result. The control unit 11 causes the display unit 13 to display the diagnostic prediction result.

As described above, according to the second embodiment, dynamic information (normal dynamic information) obtained by performing dynamic imaging of a disease-free subject managed by the medical information management apparatus 10 ) will be available later.
Specifically, machine learning is performed using the correspondence relationship between the patient parameters composed of attribute information and dynamic information accumulated in the case database 152 (see FIG. 3) and the diagnosis result, and the diagnosis target A discriminator can be generated to determine the normality or abnormality of the dynamics information. By using this learned discriminator, it is possible to predict the diagnosis result of any patient. Therefore, it is possible to prevent a doctor from overlooking a disease and assist in considering the necessity of other examinations.

In the second embodiment, it is not necessary to use attribute information as patient parameters when generating classifiers by machine learning. In this case, the control unit 11 receives dynamic information as an input and performs machine learning using a diagnosis result as an output to generate a discriminator. The control unit 11 uses a learned discriminator to input diagnostic target dynamic information, and judges whether the diagnostic target dynamic information is normal or abnormal.

[Third embodiment]
Next, a third embodiment to which the present invention is applied will be described.
The medical information management system according to the third embodiment has the same configuration as the medical information management system 100 shown in the first embodiment, so illustration and description of the configuration are omitted. The configuration and processing that are characteristic of the third embodiment will be described below.

Also in the third embodiment, the medical information management apparatus 10 utilizes AI to obtain second dynamic information (diagnostic target dynamics information) is normal or abnormal.

The control unit 11 of the medical information management apparatus 10 only obtains dynamic information (normal dynamic information) obtained by performing dynamic imaging of a subject that does not include disease and attribute information (normal attribute information) of the subject that does not include disease. Based on this, a normal model is constructed by machine learning (support vector machine, random forest, deep learning, etc.). The construction of the normal model by machine learning and the use of the normal model are realized by software processing in cooperation with the CPU of the control unit 11 and the medical information management program 151 stored in the storage unit 15 .

As patient parameters used for machine learning, attribute information (age, sex, smoking history, height, weight, BMI, respiratory function test results, etc.) and dynamic information (lung area change rate) related to normal cases (subjects not including disease) , airway diameter constriction rate, diaphragmatic velocity, etc.).

In the third embodiment, a normal model is created using past normal case data, and the degree of deviation from the normal model is calculated to determine the normality or abnormality of diagnostic target dynamic information. It is assumed that the case database 152 stores only normal case data. That is, since all the case data stored in the case database 152 do not contain disease (normal), the case database 152 does not require diagnostic results (normal/abnormal flags, diagnosis names).

When focusing only on normal patients, each parameter included in the patient parameters is considered to be intricately related, and the following characteristics can be extracted.
・Parameters that do not change consistently ・Parameters that are inconsistent and have varying values ・Parameters that are correlated (positively or negatively) with other parameters

FIG. 13 shows processing for creating a normal model by machine learning using patient parameters (normal attribute information, normal dynamics information) of normal case data.
The control unit 11 of the medical information management apparatus 10 uses machine learning to create a "normal model" by extracting features of patient parameters in normal cases. The control unit 11 automatically derives normal feature items that characterize the "normal model". For example, the control unit 11 sets the item indicating the relationship between height and BMI as the normal feature item 1, the item indicating the relationship between the smoking history and the respiratory function test result as the normal feature item 3. , search for items that show the relationship between airway stenosis rate and pulmonary function test results.

The control unit 11 calculates the degree of deviation from the "normal model" for the patient parameters (attribute information and dynamic information) related to the patient to be diagnosed, thereby determining the normality and abnormality of the dynamic information to be diagnosed.
For example, the control unit 11 calculates, for each of the normal feature items 1, 2, 3, . Then, the control unit 11 may judge an abnormality when even one of the deviance degrees corresponding to the normal feature items is larger than a predetermined threshold, or may comprehensively determine the normal feature items. It may be determined that the total deviation is recalculated, and if the total deviation is greater than a predetermined threshold value, it may be determined that there is an abnormality. There are no particular limitations on the method of determining whether the device is normal or abnormal.

In addition, the control unit 11 causes the display unit 13 to display the judgment result indicating whether or not the patient parameters related to the patient to be diagnosed are normal, the degree of deviation from the "normal model", and the like.

As described above, according to the third embodiment, dynamic information (normal dynamic information) obtained by performing dynamic imaging of a disease-free subject managed by the medical information management apparatus 10 ) can be used to calculate the degree of deviation from the normal model created, the normality or abnormality of the diagnostic target dynamic information can be determined.

In addition, based on normal case data, AI analyzes and learns the interrelationships of various normal feature items, and directly calculates the degree of deviation from the normal model (one value) as an output for the patient parameter to be diagnosed. You can do it.

In addition, in the third embodiment, attribute information may not be used as patient parameters when creating a normal model by machine learning. In this case, the control unit 11 performs machine learning using only the dynamic information of the normal case data as a patient parameter to create a normal model. The control unit 11 determines the normality or abnormality of the diagnostic target dynamic information based on the degree of deviation of the diagnostic target dynamic information from the normal model.

[Fourth embodiment]
Next, a fourth embodiment to which the present invention is applied will be described.
The medical information management system according to the fourth embodiment has the same configuration as the medical information management system 100 shown in the first embodiment, so illustration and description of the configuration are omitted. The configuration and processing characteristic of the fourth embodiment will be described below.

The medical information management apparatus 10 uses AI to determine the abnormality of second dynamic information (diagnostic target dynamic information) obtained by performing dynamic imaging of a second subject (diagnosis target patient). Present the judgment criteria when doing so.

In the fourth embodiment, the control unit 11 of the medical information management apparatus 10 uses machine learning (support vector machine, random forest, deep learning, etc.), including criteria for determining whether or not there is an abnormality. By doing so, we will learn how to judge anomalies. In the fourth embodiment, the abnormality judgment criteria (what points should be focused on) obtained by machine learning based on case data are clearly indicated to the user, and the abnormality judgment criteria are proposed. It is an example of how to use the case database 152 . The learning of the abnormality determination method, the presentation of the abnormality determination criteria, and the use of the abnormality determination method are realized by software processing in cooperation with the CPU of the control unit 11 and the medical information management program 151 stored in the storage unit 15 .

As patient parameters used for machine learning, attribute information related to each case (age, gender, smoking history, height, weight, BMI, respiratory function test results, etc.) and dynamic information (lung area change rate, airway diameter narrowing rate, diaphragm speed, etc.) are collected and stored in the case database 152 .

FIG. 14 shows a process of extracting features considered important in determining whether there is an abnormality by machine learning using patient parameters of case data.
The input data may be normal case data only, abnormal case data only, or mixed normal/abnormal case data.

When only normal case data is used as input data, the control unit 11 performs machine learning using the patient parameters of the normal case data, and from the characteristics of each parameter in the normal case, the characteristics that are important for abnormality determination (abnormality Judgment feature items) are extracted.
When only abnormal case data is used as input data, the control unit 11 extracts features that are important for abnormality judgment from the characteristics of each parameter in the abnormal case by machine learning using the patient parameters of the abnormal case data. do.
When using normal case data and abnormal case data as input data, the control unit 11 also inputs normal/abnormal labels (normal/abnormal flags) for each case data, and uses the patient parameters of each case data. Machine learning is used to extract features that are important for abnormality determination from the features of each parameter in normal cases and the features of each parameter in abnormal cases.

The control unit 11 uses machine learning to automatically derive the abnormality determination feature item. For example, the control unit 11 sets an item indicating the relationship between height and BMI as the abnormality determination feature item 1, and an item indicating the relationship between the smoking history and the result of the respiratory function test as the abnormality determination feature item 2. As item 3, find an item that indicates the airway diameter constriction rate alone.

The control unit 11 causes the display unit 13 to display the abnormality determination feature items 1, 2, 3, .

As described above, according to the fourth embodiment, by analyzing the case data managed by the medical information management apparatus 10, it is possible to propose to the user abnormality determination feature items that serve as criteria for determining abnormality. can be done.

In the fourth embodiment, it is not necessary to use attribute information as a patient parameter when extracting abnormality determination criteria and performing machine learning of an abnormality determination method. In this case, the control unit 11 performs machine learning using only the dynamic information of the case data as the patient parameter, and learns the abnormality determination method including the determination criteria.

The descriptions in each of the above embodiments are examples of the medical information management apparatus, medical information data structure, and medical information management program according to the present invention, and are not limited to these. The detailed configuration and detailed operation of each part constituting each part can also be changed as appropriate without departing from the scope of the present invention.
For example, characteristic configurations and processes in each embodiment may be combined.

Moreover, the data structure of the case database 152 is not limited to the illustrated example, and can be changed according to the application. The case database 152 may be managed for each patient or for each disease (including cases without disease).

In addition, except for the third embodiment, both normal cases and abnormal cases are accumulated in the case database 152, and the normal/abnormal cases are distinguished by the normal/abnormal flag. The case data may be accumulated separately from the database. In this case, the normal case database manages the normal dynamics information and the normal attribute information in association with each other, and the abnormal case database manages the abnormal dynamics information and the abnormal attribute information in association with each other. Furthermore, in the abnormal case database, each case data may be associated with a diagnosis name.

Further, in the case database 152, the diagnosis result for each case data may not include the diagnosis name but include only the normal/abnormal flag. In this case, the presence or absence of disease (normal or abnormal) for each case data can be used.

Even if each case data in the case database 152 does not directly include information indicating normality/abnormality (normality/abnormality flag), the presence or absence of a disease (normal or abnormal) can be determined from other information such as a diagnosis name. It may be determined. Specifically, in the case database 152, the diagnostic result for each case data does not include the normal/abnormal flag, but includes only the diagnosis name. In this case, it is possible to determine that the subject does not have any disease by leaving the diagnosis name blank or "no disease" for normal cases.

Also, a program for executing each process in each device may be stored in a portable recording medium. Also, a carrier wave may be applied as a medium for providing program data via a communication line.

10 medical information management device 11 control unit 12 operation unit 13 display unit 14 communication unit 15 storage unit 20 imaging console 30 imaging device 100 medical information management system 131 search result screen 132 analysis result screen 133 analysis result screen 151 medical information management program 152 Case database N Communication network

The present invention relates to a medical information management device and a medical information management program.

In order to solve the above-mentioned problems, the invention according to claim 1 provides: (a) movement information of the tissue obtained by performing radiographic dynamic imaging of the tissue of a subject , wherein the tissue has a disease related to the tissue; (b) management means for managing a plurality of pieces of case data associated with the subject attribute information; and statistical data of the movement information based on the plurality of pieces of case data . and output means for outputting , wherein the statistical data of the movement information is a histogram indicating the number of cases for classes obtained by classifying the movement information into a plurality of classes .

The invention according to claim 2 is characterized in that, in the medical information management apparatus according to claim 1 , the tissue is at least one of respiratory, circulatory, orthopedic, and swallowing-related tissues. and

The invention according to claim 3 is the medical information management apparatus according to claim 1 or 2 , wherein the movement information is information obtained by quantifying the movement of the tissue from the dynamic image obtained by the dynamic imaging. ,

The invention according to claim 4 is the medical information management apparatus according to any one of claims 1 to 3 , characterized in that the attribute information does not include information that identifies an individual.

The invention according to claim 5 is the medical information management apparatus according to any one of claims 1 to 4 , wherein dynamic radiation imaging is performed on the tissue of the patient to be diagnosed based on the movement information. and determining means for determining the normality or abnormality of the movement information of the patient to be diagnosed obtained in .

The invention according to claim 6 is the medical information management apparatus according to claim 5 , wherein the judgment means determines the normality or abnormality of the movement information of the patient to be diagnosed based on the movement information and the attribute information. Judging.

The invention according to claim 7 is the medical information management apparatus according to claim 5 or 6 , wherein the management means is: (d) managing a plurality of pieces of case data in which the motion information of the tissue is associated with the motion information of the third subject having a disease related to the tissue , and (d) attribute information of the third subject; characterized by

The invention according to claim 8 is the medical information management apparatus according to claim 7 , wherein the output means outputs statistics of movement information of the third subject based on a plurality of case data of the third subject. Data is output, and the statistical data of the movement information of the third subject is a histogram indicating the number of cases for classes obtained by classifying the movement information of the third subject into a plurality of classes .

The invention according to claim 9 is the medical information management apparatus according to claim 8 , wherein the output means outputs statistical data of the motion information, statistical data of the motion information of the third subject , and the diagnostic object . and outputting patient movement information.

The invention according to claim 10 is the medical information management apparatus according to claim 9 , wherein the output means outputs statistical data of the motion information, statistical data of the motion information of the third subject , and the diagnostic object . It is characterized by outputting patient movement information on the same screen.

The invention according to claim 11 is the medical information management apparatus according to any one of claims 7 to 10 , wherein the determination means is configured to perform the diagnosis based on the motion information and the motion information of the third subject . It is characterized by judging the normality or abnormality of the movement information of the target patient .

The invention according to claim 12 is the medical information management apparatus according to claim 11 , wherein the determination means comprises the movement information, the attribute information, the movement information of the third subject , and the attribute of the third subject. It is characterized by judging whether the movement information of the patient to be diagnosed is normal or abnormal based on the information.
The invention according to claim 13 is the medical information management apparatus according to any one of claims 7 to 12, characterized in that the attribute information of the third subject does not include information identifying an individual. and

The invention according to claim 14 is the medical information management apparatus according to any one of claims 5 to 13 , wherein the determination means determines the normality or abnormality of the movement information of the patient to be diagnosed by statistical analysis. Judging.

A fifteenth aspect of the invention is directed to the medical information management apparatus according to any one of the fifth to thirteenth aspects, wherein the determination means determines the normality or abnormality of the movement information of the patient to be diagnosed by machine learning. Judging.

The invention according to claim 16 is the medical information management apparatus according to any one of claims 5 to 15 , wherein the normality includes the degree of normality of the movement information of the patient to be diagnosed , and the abnormality includes the degree of abnormality in the motion information of the patient to be diagnosed .

The invention according to claim 17 provides a computer with: (a) motion information of said tissue obtained by performing dynamic radiography on said tissue of a subject , wherein said tissue does not have a disease related to said tissue; (b) management processing for managing a plurality of pieces of case data in which movement information of a subject is associated with attribute information of the subject; and output processing for outputting statistical data of the movement information based on the plurality of pieces of case data. , and the statistical data of the movement information is a histogram showing the number of cases for classes obtained by classifying the movement information into a plurality of classes .

The invention according to claim 18 is characterized in that, in the medical information management program according to claim 17 , the tissue is at least one of respiratory, circulatory, orthopedic, and swallowing related tissue. and

The invention according to claim 19 is the medical information management program according to claim 17 or 18 , wherein the movement information is information obtained by quantifying the movement of the tissue from the dynamic image obtained by the dynamic imaging. ,

The invention according to claim 20 is characterized in that, in the medical information management program according to any one of claims 17 to 19 , the attribute information does not include information that identifies an individual.

The invention according to claim 21 is the medical information management program according to any one of claims 17 to 20 , wherein the computer, based on the movement information, performs radiation dynamics on the tissue of the patient to be diagnosed . It is characterized by executing a judgment process for judging the normality or abnormality of the motion information of the patient to be diagnosed obtained by imaging.

According to a twenty -second aspect of the invention, in the medical information management program according to the twenty -first aspect, in the judgment processing, the normality or abnormality of the motion information of the patient to be diagnosed is determined based on the motion information and the attribute information. Judging.

The invention according to claim 23 is the medical information management program according to claim 21 or 22 , wherein in the management process, (d) managing a plurality of pieces of case data in which the motion information of the tissue is associated with the motion information of the third subject having a disease related to the tissue , and (d) attribute information of the third subject; characterized by

The invention according to claim 24 is the medical information management program according to claim 23 , wherein in the output processing, statistics of movement information of the third subject are calculated based on a plurality of case data of the third subject. Data is output, and the statistical data of the movement information of the third subject is a histogram indicating the number of cases for classes obtained by classifying the movement information of the third subject into a plurality of classes .

The invention according to claim 25 is the medical information management program according to claim 24 , wherein in the output process, the statistical data of the motion information, the statistical data of the motion information of the third subject , and the diagnostic object and outputting patient movement information.

The invention according to claim 26 is the medical information management program according to claim 25 , wherein in the output process, the statistical data of the motion information, the statistical data of the motion information of the third subject , and the diagnostic object . It is characterized by outputting patient movement information on the same screen.

A twenty -seventh aspect of the invention is directed to the medical information management program according to any one of the twenty- third to twenty-sixth aspects, wherein in the determination process, the diagnosis is performed based on the movement information and the movement information of the third subject . It is characterized by judging the normality or abnormality of the movement information of the target patient .

The invention according to claim 28 is the medical information management program according to claim 27 , wherein in the determination process, the movement information, the attribute information, the movement information of the third subject , and the attribute of the third subject. It is characterized by judging whether the movement information of the patient to be diagnosed is normal or abnormal based on the information.
The invention according to claim 29 is the medical information management program according to any one of claims 23 to 28, characterized in that the attribute information of the third subject does not include information identifying an individual. and

The invention according to claim 30 is the medical information management program according to any one of claims 21 to 29 , in which the normality or abnormality of the motion information of the patient to be diagnosed is determined by statistical analysis in the determination process. Judging.

The invention according to claim 31 is the medical information management program according to any one of claims 21 to 29 , wherein in the judgment processing, the normality or abnormality of the movement information of the patient to be diagnosed is determined by machine learning. Judging.

The invention according to claim 32 is the medical information management program according to any one of claims 21 to 31 , wherein the normality includes the degree of normality of the motion information of the patient to be diagnosed , and the abnormality includes the degree of abnormality in the motion information of the patient to be diagnosed .

Hereinafter, embodiments of a medical information management apparatus and a medical information management program according to the present invention will be described with reference to the drawings. It should be noted that the present invention is not limited to the illustrated examples.

The descriptions in each of the above embodiments are examples of the medical information management device and the medical information management program according to the present invention, and are not limited to these. The detailed configuration and detailed operation of each unit that constitutes each device can also be changed as appropriate without departing from the scope of the present invention.
For example, characteristic configurations and processes in each embodiment may be combined.

Claims (35)

Having management means for managing dynamic information obtained by performing dynamic imaging of a disease-free subject in association with attribute information of the subject;
A medical information management device characterized by: the disease-free subject is a subject that does not have a disease of a specific tissue;
The medical information management device according to claim 1, characterized by: the specific tissue is at least one of respiratory, circulatory, orthopedic, and swallowing-related tissues;
The medical information management device according to claim 2, characterized by: The dynamic information includes motion information obtained by analyzing dynamic images;
The medical information management device according to any one of claims 1 to 3, characterized by: the motion information is information relating to the motion of a specific tissue of the subject;
The medical information management device according to claim 4, characterized by: that the attribute information does not contain information that identifies an individual;
The medical information management device according to any one of claims 1 to 5, characterized by: having determination means for determining normality or abnormality of second dynamic information obtained by performing dynamic imaging of a second subject based on the dynamic information;
The medical information management device according to any one of claims 1 to 6, characterized by: the determining means determining normality or abnormality of the second dynamic information based on the dynamic information and the attribute information;
The medical information management device according to claim 7, characterized by: The management means associates and manages third dynamic information obtained by performing dynamic imaging of a third subject including a disease and attribute information of the third subject;
The medical information management device according to any one of claims 1 to 6, characterized by: the attribute information of the third subject does not include information that identifies an individual;
The medical information management device according to claim 9, characterized by: having output means for outputting the statistical data of the dynamic information, the statistical data of the third dynamic information, and the second dynamic information obtained by performing dynamic imaging of the second subject;
11. The medical information management device according to claim 9 or 10, characterized by: The output means outputs the statistical data of the dynamic information, the statistical data of the third dynamic information and the second dynamic information on the same screen;
The medical information management device according to claim 11, characterized by: having determination means for determining normality or abnormality of second dynamic information obtained by performing dynamic imaging of a second subject based on the dynamic information and the third dynamic information;
The medical information management device according to any one of claims 9 to 12, characterized by: the determination means determining normality or abnormality of the second dynamic information based on the dynamic information, the attribute information, the third dynamic information, and the attribute information of the third subject;
The medical information management device according to claim 13, characterized by: the determining means determining normality or abnormality of the second dynamic information by statistical analysis;
The medical information management apparatus according to any one of claims 7, 8, 13 and 14, characterized by: the determination means determining normality or abnormality of the second dynamic information by machine learning;
The medical information management apparatus according to any one of claims 7, 8, 13 and 14, characterized by: The normality includes the degree of normality of the second dynamic information,
the abnormality includes the degree of abnormality of the second dynamic information;
The medical information management apparatus according to any one of claims 7, 8, 13 and 14, characterized by: A data structure of medical information used in a dynamic information processing device that processes dynamic information obtained by performing dynamic imaging,
Dynamic information data obtained by performing dynamic imaging of a subject that does not contain a disease;
attribute information data of the subject associated with the dynamic information data;
including
A data structure for medical information characterized by: to the computer,
Executing a management process of associating and managing dynamic information obtained by performing dynamic imaging of a disease-free subject with attribute information of the subject;
A medical information management program characterized by: the disease-free subject is a subject that does not have a disease of a specific tissue;
The medical information management program according to claim 19, characterized by: the specific tissue is at least one of respiratory, circulatory, orthopedic, and swallowing-related tissues;
21. The medical information management program according to claim 20, characterized by: The dynamic information includes motion information obtained by analyzing dynamic images;
The medical information management program according to any one of claims 19 to 21, characterized by: the motion information is information relating to the motion of a specific tissue of the subject;
23. The medical information management program according to claim 22, characterized by: that the attribute information does not contain information that identifies an individual;
The medical information management program according to any one of claims 19 to 23, characterized by: on said computer;
Executing a judgment process for judging the normality or abnormality of the second dynamic information obtained by performing the dynamic imaging of the second subject based on the dynamic information;
The medical information management program according to any one of claims 19 to 24, characterized by: determining whether the second dynamic information is normal or abnormal based on the dynamic information and the attribute information in the determination process;
26. The medical information management program according to claim 25, characterized by: In the management process, third dynamic information obtained by performing dynamic imaging of a third subject including a disease and attribute information of the third subject are associated and managed;
The medical information management program according to any one of claims 19 to 24, characterized by: the attribute information of the third subject does not include information that identifies an individual;
28. The medical information management program according to claim 27, characterized by: on said computer;
executing output processing for outputting the statistical data of the dynamic information, the statistical data of the third dynamic information, and the second dynamic information obtained by performing dynamic imaging of a second subject;
29. The medical information management program according to claim 27 or 28, characterized by: In the output process, the statistical data of the dynamic information, the statistical data of the third dynamic information, and the second dynamic information are output on the same screen;
30. The medical information management program according to claim 29, characterized by: on said computer;
executing a judgment process for judging the normality or abnormality of the second dynamic information obtained by performing the dynamic imaging of the second subject based on the dynamic information and the third dynamic information;
The medical information management program according to any one of claims 27 to 30, characterized by: determining whether the second dynamic information is normal or abnormal based on the dynamic information, the attribute information, the third dynamic information, and the attribute information of the third subject in the determination process;
32. The medical information management program according to claim 31, characterized by: In the determination process, determining normality or abnormality of the second dynamic information by statistical analysis;
33. The medical information management program according to any one of claims 25, 26, 31 and 32, characterized by: determining the normality or abnormality of the second dynamic information by machine learning in the determination process;
33. The medical information management program according to any one of claims 25, 26, 31 and 32, characterized by: The normality includes the degree of normality of the second dynamic information,
the abnormality includes the degree of abnormality of the second dynamic information;
33. The medical information management program according to any one of claims 25, 26, 31 and 32, characterized by:

Images